US 5623596 A
A power system fault handling mechanism for portable electronic devices having multiple power supplies. A safety-net circuit is triggered when a power fault condition is detected. The safety-net circuit is a diode-based bridging circuit which couples all available power supplies to the power load, yet avoids feeding power back from any of the supplies to one another. This ensures that the load continues to receive power after a fault condition occurs without interruption. A microcontroller may then implement a routine to switch to a working power supply so that the device may resume normal operation. The power system fault detection mechanism therefor allows for the switching between system power supplies without disrupting a running process.
1. A computer system comprising:
a processing unit;
a memory bus coupled to the processing unit, the memory bus conveys instructions and data to and from the processing unit;
a memory subsystem coupled to the memory bus, the memory subsystem contains the instructions and data; and
a power system which proyides operating power to the computer system, the power system including:
a plurality of available power supplies which supply an operating voltage to the computer system, the plurality of available power supplies includes
an AC adapter selectively chosen by the power system for providing the operating voltage to the computer system,
a first battery selectively chosen by the power system for providing the operating voltage to the computer system, and
a second battery selectively chosen by the power system for providing the operating voltage to the computer system,
wherein during normal operation of the computer system only one of the plurality of power supplies is selectively chosen by the power system for providing the operating voltage to the computer system,
power fault detection circuitry to detect a power system fault condition, and
power supply bridging circuitry, responsive to the power fault detection circuitry, to operatively couple the plurality of available power supplies together to supply power to the computer system when said power system fault condition exists.
2. The computer system of claim 1 wherein the power supply bridging circuitry comprises:
a first power path coupled between the AC adapter and the computer system, the first power path delivers power from the AC adapter to the computer system;
a second power path coupled between the first battery and the computer system and selectively coupled by the power system to the first power path;
a third power path coupled between the second battery and the computer system and selectively coupled by the power system to the first and second power paths; and
a power supply bridging switch to electrically couple the first, second and third power paths when a power fault condition is detected by the power fault detection circuitry, the power supply bridging switch responsive to the power fault detection circuitry.
3. The computer system of claim 2 wherein the power supply bridging circuitry further comprises:
a first diode implemented in the first power path and oriented to prevent power from flowing toward the AC adapter;
a second diode implemented in the second power path and oriented to prevent power from flowing toward the first battery; and
a third diode implemented in the third power path and oriented to prevent power from flowing toward the second battery.
4. The computer system of claim 3 wherein the power fault detection circuitry comprises a first comparator which compares the operating voltage with a threshold voltage level, the first comparator generating a fault detection signal when the operating voltage falls below the threshold voltage level.
5. The computer system of claim 4 further comprising:
power system fault recovery logic responsive to the power supply fault detection circuitry for recovering from the power system fault condition and selecting one of the plurality of available power supplies to provide an operating voltage to the computer system.
6. The computer system of claim 5 wherein the power system fault recovery logic receives an applied logic voltage, the power system further comprising a second power supply fault detection circuitry which detects a power fault condition relating to the applied logic voltage.
7. The computer system of claim 6 wherein the second power supply fault detection circuitry comprises a second comparator which compares the applied logic voltage to a threshold voltage, the second comparator propagating a second power fault detection signal when the applied logic voltage falls below the threshold voltage.
8. The computer system of claim 7 further comprising a backup linear regulator which supplies power to the power system fault recovery logic during a fault condition, the backup linear regulator being activated by the power supply fault detection circuitry.
1. Related Applications
The present invention is related to co-pending U.S. patent application entitled "Power System Configuration and Recovery from a Power Fault Condition in a Computer System Having Multiple Power Supplies", Ser. No. 08/239,602, assigned to the assignee of the present invention and filed concurrently herewith.
2. Field of the Invention
The present invention relates to power management for portable electronic systems. More particularly, the present invention relates to power fault detection and protection methods and mechanisms for portable computer systems having multiple power sources.
3. Art Background
As the computer revolution has progressed, computing machines have become ever smaller, yet more powerful. The advent of advanced semiconductor technologies has resulted in computers that are easily transportable. Today, portable laptop computers which utilize advanced packaging techniques are nearly as common as desktop computers were just a few years ago. The efforts of computer architects today, using emerging technologies, is the ever smaller, faster portable computer that is lightweight, extremely modular and flexible in its use.
An important consideration in designing portable computing devices, or any portable electronic device is the configuration of the power system. Ideally, when a computer is used at a fixed location, the user should be able to plug the device in to an AC-power outlet and derive power for the device by means of an AC adapter or equivalent technology, where DC power is supplied to the computer after being converted from AC by the AC adapter. When the portability aspect of the device is being utilized, and electric wall sockets are unavailable, the portable computer must operate from battery power. While battery technologies are improving, they still suffer the unfortunate characteristics of being large and heavy relative to the overall size and weight of portable devices.
In an effort to meet the above goals of lightweight, modular systems having increased flexibility, computer makers have begun exploring the possibility of providing multiple power sources, including more than one battery compartment. This allows for more flexibility in designing around space constraints in tightly packaged small systems. The price of this increased flexibility is increased complexity with respect to power management subsystems. For example, when more than one battery pack is provided for, the computer system must provide additional battery contacts and provide for the coordination of selecting an optimal operating power source.
Another object of the evolving computer revolution has been to make computers more "friendly". Computers that lose data or require re-initializing because of power problems are decidedly unfriendly. Further, it is a disadvantage of many systems to have to shut down the computer system when changing from one power supply to another. It would be desirable, and is therefore an object of the present invention, to provide a reliable power supply management system for portable computers having multiple power sources.
From the foregoing, it can be appreciated that there is a need to provide superior power system control techniques for densely integrated portable electronic systems having multiple power supplies. Accordingly, it is an object of the present invention to provide a method and apparatus for the seamless transition between power sources in a portable computer system having multiple power sources. It is also an object of the present invention to provide a fault detection and protection technique for a computer system that encounters a power fault when operating from one of several available power supplies. It is another object of the present invention to provide a method and apparatus to recover from such power faults in such a way that power source transitions are not disruptive to a running computer system.
These and other objects of the present invention are provided in a preferred embodiment portable computer system that utilizes four (4) available power supplies: power from an AC adapter and power cord, two separate battery power bays, and a state-preserving rechargeable lithium coin-stack battery. During normal computer system operation, only one power source provides power to the components of the computer system. The power management system is driven by a software operated microcontroller which provides driving signals to a Programmable Array Logic (PAL) device for operating the power system's switching circuits.
The power system is provided with fault detection circuitry which monitors the voltage level of the power being supplied to the computer system components. If that supply voltage drops below a threshold level, a safety-net circuit is triggered. The threshold level for activating the safety-net circuit varies depending on whether the system is being powered by an AC line voltage or by a battery.
The safety-net circuit that is triggered upon detection of a power fault provides a fault indicating signal to the PAL. When the fault indicating signal is received by the PAL, it propagates appropriate signals to a plurality of the power system's switches which cause all available power supplies to be bridged through an arrangement of diodes. This guarantees that the load (the computer system) will continue to receive power from the combined power supply available to the system. Additionally, while the fault condition is active, a backup low power regulator is provided with a control signal which turns the regulator on for ensuring that the system's parameter memory and the microcontroller are maintained in a safe state.
Once a power system fault condition has been detected and the available power supplies are bridged by the safety-net circuit, a routine running in the power system microcontroller may begin the process of determining the best way to recover from the power system fault condition in a controlled fashion (i.e., selecting the appropriate power supply to designate as the system's operating power supply). Once a power supply is designated as the system's power supply, the fault condition transition situation is terminated and the computer system resumes normal operation. The entire process may occur without interruption to a user operating the computer system. This mechanism allows a user to switch from an AC power source to a battery source without halting a running process. Likewise, a user may remove one battery pack and continue operating from the other without having to shut down the system.
The objects, features and advantages of the present invention will be apparent from the following detailed description in which:
FIG. 1 illustrates a block diagram of a computer architecture incorporating the present invention power system fault handling mechanism.
FIG. 2 illustrates the various states in which a portable computer incorporating the present invention may operate.
FIG. 3 illustrates a circuit diagram of a computer power control system incorporating the present invention.
FIG. 4 illustrates a more detailed circuit diagram of the present invention power supply fault detection mechanism.
FIG. 5 illustrates a general flow diagram for recovering from a power system fault detected by the present invention.
A method and apparatus are described for the handling of power system fault conditions in one embodiment of a portable computer system which utilizes two batteries as well as having the ability to operate from an AC line voltage via an AC adapter. The implementation of the present invention in a particular electronic device should not be seen as limiting because the present invention may be suitable for use in any portable electronic device capable of operating from multiple available power sources. Throughout this detailed description numerous details are specified such as nominal voltages and discrete signal names, in order to provide a thorough understanding of the present invention. To one skilled in the art, however, it will be appreciated that the present invention may be practiced without such specific details. In other instances, well-known control structures and gate level circuits have not been shown in detail in order not to obscure the present invention. Particularly, a microcontroller and separate programmable array logic (PAL) are described for implementing the control circuitry in one embodiment.
With today's device technology, the development of specialized integrated circuits and programmable logic generally do not require the rendering of fully detailed circuit diagrams. The definition of logic functionality and state diagrams allow computer design techniques to design the desired logic and circuits. Additionally, microcontrollers are known to operate based on a desired flow diagram rendered into software that is compatible with a selected microcontroller. Accordingly, portions of the present invention will be described primarily in terms of functionality to be implemented by a microcontroller and a PAL. This functionality will be described in detail with logic equations and flow diagrams will be used for illustration. Those of ordinary skill in the art, once given the following descriptions of the various functions to be carried out by the present invention will be able to implement the necessary microcontroller structure and logic for the PAL in suitable technologies without undue experimentation. As an alternative to using a PAL, those of ordinary skill in the art will recognize that various random logic devices or a custom designed integrated circuit could be implemented.
A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent disclosure, as it appears in the Patent and Trademark Office patent files or records, but otherwise reserves all copyrights whatsoever.
Referring now to FIG. 1, one embodiment of a portable computer architecture implementing the present invention is shown. The constituents of the computer 5 are shown coupled to a CPU or memory bus 15. The memory bus 15 provides the signal path for the exchanging of signals between the various elements coupled to the memory bus. Further provided by the memory bus are control lines for such things as bus request and bus granting signals and other system control signals. As with many conventional computer systems, the architecture illustrated in FIG. 1 includes a CPU 10, ROM 12 and the system's main memory subsystem 14 coupled to the memory bus 15. It should be understood that in the illustration the main memory subsystem 14 coupled to the memory bus refers to the system's DRAM and the controller required for writing to and reading from the DRAM based on a requested transaction.
Also shown coupled to the memory bus 15 in the computer 5 is an I/O interface 20 which is used for coupling the memory bus 15 to the I/O bus 30 for providing the computer system with a number of I/O capabilities. The I/O interface 20 shown in the figure incorporates bus interface logic 21 which controls the transactions between the main memory system 14 and I/O devices. The I/O bus 30 provides the avenue for a number of different I/O devices to be incorporated in the computer system. For example, the SCSI port 31 is coupled to the I/O bus 30. SCSI stands for Small Computer System Interface which is one of the computer industry standards for coupling I/O devices such as the hard disk drive 32 or floppy drives or other storage media and other I/O devices. The computer system also has a serial port 35 which is shown coupled to the I/O bus 30. The serial port 35 can be used for attaching a modem or printer or other serial devices. Additionally, the I/O bus 30 may be used to couple the computer system to a local area network (LAN) 39 through an ethernet port 38. This allows the computer system to communicate with other computer systems and share common resources such as community printers, etc. In portable computer systems, an ethernet port provides for a rapid attachment and separation of the portable system to a running network. Finally, there is shown coupled to the I/O bus 30 a keyboard 33 and display 34 which provide the user's interface to the computer 5. Other devices, such as a floppy disk drive may also be incorporated into the computer 5.
Most of the computer system components shown in FIG. 1 require electrical power in order to operate. Therefore, in FIG. 1 there is shown coupled to the computer 5 the power system 100. The power system 100 is the focus of the present invention which is tailored for a portable computing device. One embodiment of the present invention power system provides for an external AC adapter to provide power when AC line voltage is available. The power system also provides for two separate battery packs to be utilized in operating the computer when it is being used in its portable capacity. Finally, a backup battery pack is provided for maintaining the computer system's state when no other power sources are available. The operational relationship of these power supplies will be described below.
Desktop computers generally exist in one of two power states: on or off. With battery operated computers, for power savings purposes, a third state is introduced. FIG. 2 is a state diagram illustrating three possible power states of a battery operated computer. During normal operation the system exists in a run state 301 in which all components of the computer may be active and are operationally available for performing their tasks. A sleep state 302 is introduced that the computer system may enter from the run state, generally after a predetermined period of inactivity or if instructed by the user of the computer. In the sleep state, the power consumption of the system is reduced by deactivating various energy consuming portions of the computer system that are not currently needed. The run state 301 may be re-entered from the sleep state 302, for example, by the depressing of any key on the computer system's keyboard in which case power will then go back to those components which were deactivated in the sleep state. The shut down state 303 corresponds to the usual off state for computers, but for portable computers such things as charging of batteries and other internal maintenance may occur in this state. The transition from the shut down state 303 to the run state 301 occurs if the user depresses the power-on key for the computer system and if there is sufficient power available to run the computer system. A transition directly from the run to the shut down state is available which the user may initiate when finished using the computer or which may occur if the power supply becomes insufficient to operate the computer system.
One portable computer system incorporating the present invention is configured with two power bays each for receiving battery packs. The portable computer system implements a pointing device in the lower central portion of the computer casing which is one of the reasons necessitating two power bays. The use of multiple power bays also provides for increased modularity in portable systems.
Referring now to FIG. 3 a power system oriented view of the computer system of the present invention is shown. The power system 100 is shown in detail coupled generally to the computer 5. The diagram of FIG. 3 illustrates four (4) possible power supplies for the computer system: an external AC adapter 101, two separate batteries 102 and 103 as well as a backup battery. 104. The external AC adapter 101 is used to provide power to the computer 5 through the buck regulator 105 when there is an available AC line voltage to plug the computer system into. The AC adapter optimally provides 16.8 volts output to the system and is also used for charging the batteries 102 and 103 when they require charging and the adapter is plugged-in to an available AC line voltage. Note that although the illustrated embodiment indicates separate outputs from the AC adapter (V-- BATT and V-- MAIN), a simpler AC adapter could be employed that collapses these two outputs into a single output.
The lithium backup coin-cell batteries 104 are used when no other power source is available to the system. They provide a minimal power supply to the computer system when it is operating in sleep mode or during shut down to preserve the parameter RAM of the computer system as well as other states such as monitor configurations, etc. More about the operation of the backup cells 104 will be described below.
Batteries 102 and 103 are the power supplies tapped by the system when no AC line voltage is available. In one embodiment, the batteries are rechargeable nickel-metal-hybride batteries. Having both batteries present in FIG. 3 best exemplifies the operation of the present invention. There are some circumstances in which only one battery is present for providing power to the system when no AC line voltage is available. During normal operation of the computer by battery, an intelligent battery discharging scheme is implemented for selecting which battery should provide the computer system with power. The actual battery hierarchical discharging scheme is beyond the scope of the present invention which is concerned with fault management regardless of the chosen discharging scheme. The present invention merely assumes that during normal operation, one power source will be selected according to some predetermined criteria.
Essentially, microcontroller 110 is programmed with a routine which optimizes the system for longevity and reliability. The microcontroller 110 provides operating signals to the programmable array logic (PAL) 115 which provides the actual control signals to the components of the power management system. For example, when the computer system is operating from an AC line voltage through AC adapter 101, the microcontroller 110 may determine that either battery 102 or battery 103 (or both) should be recharged by power from the AC adapter 101. For the charging of battery 102, the microcontroller sends a signal PGE-- ICHG-- 1-- L to the PAL 115. The PAL 115 in response to the PGE-- ICHG-- 1-- L signal propagates a CHARGE-- 1 signal to the CHARGE-- 1 FET 116. Similarly, if the microcontroller 110 determines that the battery 103 should be charged from the AC adapter 101, the microcontroller propagates a PGE-- ICHG-- 2-- L to the PAL 115 which in turn activates the CHARGE-- 2 signal to the gate of the CHARGE-- 2 FET 117. Note also that for the microcontroller 110 to instruct the system to run from power supplied by the AC adapter 101, it propagates a signal PGE-- VCHG-- ON-- L to the PAL 115 which in turn provides the CHGR-- ON signal to the CHGR-- ON FET 118. The activation of the FET 118 provides a power path from the AC adapter 101 for the operating voltage through the buck regulator 105 to the rest of the computer system 5.
When the computer is not plugged into an AC line voltage, the microcontroller 110 makes a determination as to which battery should be operating the system according to the discharging rules under which it operates. If it determines that battery 102 should be utilized, then the microcontroller provides PGE-- LOAD-- 1 signal to the PAL 115 which responsively sends the signal LOAD-- 1 to the LOAD-- 1 FET 119. Similarly, if the microcontroller 110 determines that the battery 103 should be used to power the system, then the microcontroller provides a PGE-- LOAD-- 2 control to the PAL 115 which in turn propagates the LOAD-- 2 signal to the LOAD-- 2 FET 120.
The power being provided from the selected power source to the system is identified in the figure has V-- 18P-- MAIN. This source is also provided to a fault detection comparator 125. In addition to the V-- 18P-- MAIN source being provided to the fault comparator 125, a signal designated FAULT-- LVL is also input to the comparator 125. When the V-- 18P-- MAIN signal drops below the FAULT-- LVL threshold, a fault condition is considered to exist. The comparator 125 then provides an active FAULT-- PAL signal to the PAL 115.
The source of the FAULT-- LVL threshold signal is from the PAL 115. The fault level threshold varies depending on which power source the system is running from. If the AC adapter 101 is being used to power the system, the fault level is set to approximately 15.5 volts which is about 1.3 volts lower than the 16.8 volt normal operating voltage from the AC adapter 101. The fault level is set to a different value if the computer system is running from one of the batteries. Batteries are subject to a wider swing in voltage ranges depending upon the charge levels. The normal range of the battery output is expected to be between 8 and 11.5 volts, with approximately 10 volts being nominal. Accordingly, when the computer system is operating from one of the batteries, the FAULT-- LVL threshold is set to approximately 7.5 volts.
When the fault detection comparator 125 detects that the V-- 18P-- MAIN signal has dropped below the fault level threshold, a resulting FAULT-- PAL signal is provided to PAL 115 which activates the output signal FAULT-- H from the PAL. The FAULT-- H signal is provided to the gate of fault detection system trigger FET 126. When FET 126 is activated, all available system power supplies are bridged together through diodes 127, 128, 129 and 130. The diodes ensure that power only flows away from the power supplies so none are shorted together. Simultaneously, FETs 116-120 are deactivated by the PAL 115 to avoid shorting power supplies together. Note that in the event that an AC adapter is implemented that has only one output line, diodes 129 and 130 would be replaced with a single diode for coupling that output from the AC adapter in the bridged circuit.
Regardless of what power supply was operating the computer system, once a fault condition is detected, all the power supplies are coupled together to provide the V-- 18P-- MAIN source to the computer system. This guarantees a power supply to the computer system during the fault state. The FAULT-- H signal is also propagated to the microcontroller 110 to inform the microcontroller that a fault condition exists and directing it to initiate its fault handling routine for selecting an appropriate single power source to operate from in a controlled fashion. The dashed path of FIG. 3 illustrates the power supply propagation to the system when a fault condition exists.
It should also be noted that the power source to the microcontroller 110 and the PAL 115 is a power line noted as V-- 5P-- PGE which is an output from the buck regulator 105 or the linear regulator 135. The buck regulator 105 powers this V-- 5P-- PGE line through FET 132 in the run and sleep states during normal operation. However, if FAULT-- H is active, the linear regulator is used to power V-- 5P-- PGE with FET 132 turned off (open). This protects the state of the microcontroller 110 and PAL 115 as a second safety net. More particularly, when a fault condition exists and the FAULT-- H signal has gone high, the FET 132 is deactivated, responsive to the BREG-- SW-- L signal from the PAL to choke off the power supply from the buck regulator 105 to the microcontroller 110 and PAL 115. In the fault condition state, the PAL 115 also outputs a signal RUN-- LIN-- L which is provided to the backup linear regulator 135. This signal activates the linear regulator 135. Accordingly, when the fault condition exists, the bridged power supply source V-- 18P-- MAIN as well as the power supply source from the backup battery 104 and AC adaptor 101 through V-- MAIN are supplied to the linear regulator 135 through bridging diodes 136, 137 and 138. The RUN-- LIN-- L signal, activating the backup linear regulator 135, causes it to output the V-- 5P-- PGE signal to the microcontroller 110 and the PAL 115 to ensure that they operate properly during the existence of a fault condition.
Referring now to FIG. 4, the circuit diagram of FIG. 3 is shown when the fault condition is active. The FETs 116, 117, 118, 119 and 120 are replaced with a symbol to indicate that those FETs are off. The FET 126 is "on" as described above, which creates a path for the power to be supplied from all sources via the path dictated by diodes 127, 128, 129 and 130. Also, the V-- 5P-- PGE supply to the microcontroller 110 and PAL 115 are provided through the backup linear regulator 135. In FIG. 4, the PAL 115 is shown to include the functionality of a collection of AND gates 142. It can be seen that when the FAULT-- PAL signal is sent active from the comparator 125, a flip-flop 131 changes state and then provides a signal to the functional equivalent of AND gates 142 to provide the signals to shut-off the FETs described above. The flip-flop 131 also allows the fault indicator signal FAULT-- H to be provided to the microcontroller 110 and to the fault condition trigger FET 126. As was described, the FAULT-- H signal is provided to the microcontroller 110 which notifies it to initiate its routines for resolving and recovering from the fault condition. Once the microcontroller 110 has completed the resolution of the fault condition, it sends a PGE-- FAULT-- CLR signal to the PAL 115 which resets the flip-flop 131 to the normal operating state and instructs the PAL which power supply should be provided an activation signal therefrom. The microcontroller 110 may wait a brief period of time before clearing the fault condition state in order to allow any transient fault conditions to resolve themselves.
Another aspect extending the present invention is the backup safety-net comparator 140. This comparator receives as one input the V-- 5P-- PGE operating voltage that is supplied to the microcontroller 110 and PAL 115. A second input is set to approximately 4.5 volts. In the event that the operating voltage for the microcontroller and PAL drops below this trigger level, the comparator 140 sends a signal to reset the microcontroller 110 to a known state and provides the same signal to the PAL 115 which is used in determining the output signals described above and which will be described below with respect to the logic equations to be implemented by the PAL 115. Essentially this signal from comparator 140 causes the outputs of PAL 115 to go to nonhazardous states. One reason for monitoring the operating voltage of the microcontroller 110 and PAL 115 is to prevent them from sending spurious signals to the system when their power supply voltage falls below the acceptable operating region required for normal operation which could affect the software operation of the microcontroller. A particular concern is that the microcontroller not send a signal instructing a shorting of the two batteries together which could be disruptive to the system and certainly hazardous to the batteries,
Table I below identifies the inputs and outputs for one embodiment of the PAL logic 115, Table II below illustrates the random logic equations to be carried out by the PAL 115 in generating its output signals based on the above identified input signals, Table III below illustrates the logic performed by the comparators in monitoring for various fault conditions,
TABLE I______________________________________PAL Logic DescriptionBuffer Logic Inputs Buffer Logic Outputs______________________________________PGE-- RUN BREG-- SW-- LPGE-- SLEEP RUN-- LIN-- LPGE-- LOAD-- 1 LOAD-- 1PGE-- LOAD-- 2 LOAD-- 2PGE-- VCHG-- ON-- L CHARGE-- 1PGE-- ICHG-- 1-- L CHARGE-- 2PGE-- ICHG-- 2-- L CHGR-- ONFAULT-- PAL FAULT-- LVLPGE-- FAULT-- CLR FAULT-- HPWR-- OK______________________________________
TABLE II__________________________________________________________________________Random Logic Equations__________________________________________________________________________FAULT-- H = FAULT-- H *!PGE-- FAULT-- CLR+FAULT-- PAL*(PGE-- RUN+PGE-- SLEEP)FAULT-- LVL = !PGE-- VCHG-- ON-- L*!PGE-- LOAD-- 1*!PGE-- LOAD-- 2*!PGE-- SLEEPLOAD-- 1 = !FAULT-- H*PGE-- LOAD-- 1*!PGE-- LOAD-- 2LOAD-- 2 = (!FAULT-- H*PGE-- LOAD-- 2*!PGE-- LOAD- - 1)+!PDS-- IN-- L*(PGE-- SLEEP+PGE-- RUN)*PWR-- OKCHARGE-- 2 = !FAULT-- H*!PGE-- ICHG-- 2-- L*PGE-- ICHG-- 1-- LBREG-- SW-- L = !(!FAULT-- PAL*PGE-- RUN*PWR-- OK+PGE-- SLEEP*FAULT-- PAL* PWR-- OK)RUN-- LIN-- L = !FAULT-- PAL*PGE-- RUN*PWR-- OKCHARGE-- 1 = !(FAULT-- H+PGE-- ICHG-- 1-- L)CHGR-- ON = !(FAULT-- H+PGE-- VCHG-- ON-- L)__________________________________________________________________________
TABLE III______________________________________Comparator Equation______________________________________FAULT-- PAL = (V-- 18P-- MAIN<REF.) REF=15.0V (charger in) or 7.5V (charger out)PWR-- OK = (V-- 5P-- PGE<4.5V)______________________________________
The convention used in the above tables with respect to the random logic equations is that `!`=NOT; and asterisk, `*`=AND, and a plus sign, `+`=OR in evaluating the Boolean equations listed. In one embodiment, the signals are active high unless they are denoted with an `-- L`. In addition to the signals already described above, some of the expressions use terms not yet identified. Note that PGE-- RUN indicates . that the system is in the run state and PGE-- SLEEP indicates that the system is in the sleep state. These states are contrasted with the shut down state as described above with respect to FIG. 2. The PDS-- IN-- L signal referred to in the LOAD-- 2 equation is a signal that is active when something other than a battery is in the second battery bay instead of battery 103. Note that derivatives of this invention could use a subset or extension of the equations above to satisfy the needs of a particular power system implementation with or without expansion capabilities. Similarly, the comparator equation voltages can be altered to suit the needs dictated by battery and/or different AC adapter voltages.
In one embodiment of the power fault detection and protection mechanism, the bridging circuitry is only activated if the computer system is operating in either the run or sleep states which were described above. According to the logic equations in Table II, the FAULT-- H signal is not allowed to go active in the event that the computer is in the shut down state. In the shut down state, however, the comparator 140 still monitors to make sure that the microcontroller 110 and the PAL 115 receive an appropriate operating voltage. It should be understood that these circuits may continue operating even in the shut down state to carry out such things as battery recharging operating a built-in real-time clock and preserving a portion of memory dedicated to monitor configurations and virtual desktops, etc.
The description of the present invention power system is one that is software controlled through the microcontroller 110. The PAL 115 serves as a buffer to protect the power system from spurious controls that might occur in a fault condition. When a fault condition is detected, the power system is taken over by the comparator and PAL 115 in the above noted logic which prevents the microcontroller from issuing invalid and potentially dangerous signal state combinations to the power system switches. More importantly, the comparator and PAL are able to react much more quickly than the microcontroller which would suffer from software latency as a function of clock frequency. Therefore, the design keeps software based delays out of the response to a fault, until the transient has passed. Otherwise, a very high clock speed microcontroller (or powerful microcontroller) would be required to respond fast enough to "save" the computer state in a fault transient condition of falling voltage. Once in the fault condition state, the microcontroller may then recover and execute the necessary routines for recovering from the fault condition.
By buffering the microcontroller signals using the PAL and the safety-net circuit described above, a computer system in accordance with the present invention is allowed to continue operating through the existence of a fault condition. This provides the computer user with the flexibility to exchange power supplies "on the fly" without having to interrupt an operating session on the computer. Additionally, unforeseen accidents such as kicking a plug out of a wall socket are not as disastrous since the system automatically slides into the fault condition state until the microcontroller has time to shift the operation of the system to one of the batteries.
Once the power control system of the present invention has entered into the fault condition state in which all power supplies are bridged together to ensure a continuing power supply to the computer system, the power system's microcontroller may initiate a fault recovery routine to resume normal operation of the computer from an appropriate single power supply. FIG. 5 illustrates one embodiment of a general flow diagram to be executed by the microcontroller in recovering from the power fault condition designated as Fault Recovery routine 200. At decision box 201, the routine determines if the machine is plugged in and if AC power is available. In general, operating from an AC voltage is the optimal choice for the power system. If so, then at step 202 the decision is made to run from the AC adapter and the routine proceeds to step 215 in which the microcontroller signals the PGE-- FAULT-- CLEAR signal to the PAL to take the power system out of the fault condition state and resume normal operation. In the case where it has been determined that the AC adapter should be the system's power supply, the microcontroller will propagate the PGE-- VCHG-- ON-- L to the PAL 115 which will in turn signal the CHGR-- ON signal to the CHGR-- ON FET 118.
If the machine is not plugged in or AC power is not available, then the routine 200 will proceed to decision box 204 from which it determines if battery 2 is available and whether it is strong enough to support the machine according to criteria defined for the microcontroller. If so, then at box 205 the state is set for the machine to operate from battery 2 and the routine proceeds to step 215 in which the fault clear signal is provided to the PAL which responsively propagates the LOAD-- 2 signal to the LOAD-- 2 FET described above. If battery 2 is not available or charged sufficiently, then at decision box 207 the routine determines whether or not battery 1 is available and whether it is strong enough to operate the system according to the criteria programmed into the microcontroller. If so, then at step 208 the microcontroller activates the signals to operate the system from battery 1 and the Fault Recovery program 200 proceeds to step 215. In this case, the microcontroller provides the fault clear signal to the PAL which responsively activates the LOAD-- 1 signal to the LOAD-- 1 FET described above to designate battery 1 as the power source for the system.
Finally, if it turns out that no AC power is available and batteries 1 and 2 are incapable of operating the system, then at step 210 the routine signals to the user that there is an unresolvable problem, designates the machine to operate in a reduced power condition of a sleep state or shut down state perhaps requiring the use of lithium backup batteries at this point, and therein concludes operation.
The above general fault recovery routine described with respect to FIG. 5 may be more elaborate and tailored to suit the needs of a particular system in which it is to be implemented. Co-pending patent application "Power System Configuration and Recovery from a Power Fault Condition in a Computer System Having Multiple Power Supplies", Ser. No. 08/239,602, assigned to the assignee of the present invention and filed concurrently herewith provides additional details for the fault recovery routine for one embodiment of a computer system as described above. That application is incorporated herein by reference.
There has thus been described a fault detection and protection mechanism for use in a computer system having multiple power supplies. The mechanism quickly intercepts fault conditions and protects the system while operating in the fault state by bridging all available power supplies. Once that state has been entered, a software driven microcontroller will then have time to determine the best route to recover from the fault. Although the present invention has been described in terms of various embodiments, it will be appreciated that various modifications and alterations may be made by those skilled in the art without departing from the spirit and scope of the invention. The invention should, therefore, be measured in terms of the claims which follow.